The present invention relates generally to the field of computer memory cache access and instruction pipelining, and more particularly to improving processor efficiency in memory caching by modifying addresses of fetch requests to maintain data return order.